SUBJECT SUBJECT LEVEL TIME / DU DATE
1. Pleas 2. This q 3. Pleas 4. Answ illustr 5. This q
quest 6. Answ 7. Graph
CODE TITLE
RATION
se read the i question pa se write you wer should b
ration.
question pa tions in Sec wer all quest h paper is a
UNIV Ma F SEPT
: : : : :
INST
instructions aper is print ur answers o
be written in
aper consis ction A. For tions in Eng appended.
ERSITI K laysia F INAL EX TEMBER
FSB 33503 IMAGE PR BACHELO 3 HOURS
TRUCTION
s given in th ted on both on the answ n blue or bl
ts of TWO ( r Section B,
glish.
KUALA France In
XAMINA R 2013 S
3
ROCESSIN OR
NS TO CAN
he question h sides of th
wer booklet ack ink exc
(2) sections answer two
LUMPU nstitute ATION
SESSION
NG
NDIDATES
n paper CAR he paper.
t provided.
cept for ske
s, Section A o (2) quest
UR
N
REFULLY.
etching, gra
A and B. Ans tions only.
phic and
swer ALL
SET A
FSB33503 / FSB43503 IMAGE PROCESSING 1
SECTION A (Total: 40 marks)
INSTRUCTION: Answer ALL questions.
Please use the answer booklet provided.
Question 1
(a) List two (2) areas of image processing application. Give an example of each application
(4 marks) (b) An image can be viewed in 2D and 3D formation. Briefly discuss the 2D and 3D
image formation
(4 marks) (c) Give two (2) broad techniques of image enhancement
(2 marks) (d) Figure 1 shows a matrix images with 5x5 scales. Calculate the mean, variance,
standard deviation and histogram of its pixel values. Show the formula for mean, variance, standard deviation and histogram as a bar plot frequency of pixel vs pixel value
1 0 3 2 3 0 0 0 1 0 1 1 2 2 4 8 0 1 4 8 8 5 5 0 2 Figure 1: Matrix image of 5x5 scales
(10 marks)
Question 2
(a) Explain briefly about image segmentation in term of meaning, goal and process in image processing .
(5 marks)
(b) An object can be easily detected in an image if the object has sufficient contrast from the background. We use edge detection and basic morphology tools to detect a prostate cancer cell. The detection requires six (6) steps. State and explain in details for each step.
(12 marks)
(c) Refer to Figure 2, state the process involved to transform Image A to Image B.
Image A Image B
Figure 2 (3 marks)
FSB33503 / FSB43503 IMAGE PROCESSING 3
SECTION B (Total: 60 marks)
INSTRUCTION: Answer TWO (2) questions only.
Please use the answer booklet provided.
Question 3
Figure 3 shows two (2) images and will be implemented extraction for classification.
Image A =
Image B =
98 5 0 2 0
10 0 0 0 9
0 0 0 10 2
0 100 0 8 70 5
199 0 5 0 1
Figure 3: Matrix images of each class
(a) Sketch the view of the matrix images.
(4 marks)
(b) The matrix images have been converted to BW format with threshold value 10. Give the pixel values of both matrix images in Black and White format.
(4 marks)
(c) Compute the mean and standard deviation of both class images?
(4 marks)
(d) Write an algorithm to classify the images by using the mean statistical properties.
(8 marks)
98 55 50 122 204
100 0 0 0 109
120 0 0 10 231
100 0 0 0 155
199 200 255 200 100
(e) Based on the two images in Question 3(a), write the matlab code for below operations:
i. Add the two images and display the output images using subplot
(4 marks)
ii. Subtract between the images using imsubstract and imabsdiff. Display the output using subplot. Justify why the image output for this two functions displaying different intensity value.
(6 marks)
Question 4
(a) Briefly explain about image enhancement and why we need to enhance the image.
(5 marks)
(b) State two technique of point processing enhancement
(4 marks)
(c) Figure 4 shows the image that has been resized to 1.25 times from the original image. Write the matlab code to display this image as shown in Figure 4. (We could assume that the filename of the grayscale image is “circuit.tif”).
(6 marks)
Figure 4: The resize image Original Image
The resize image
FSB33503 / FSB43503 IMAGE PROCESSING 5
(d) Linear filtering of an image is accomplished through an operation called convolution.
Based on the matrix image in Table 1, explain the four(4) steps to compute the (3,3) pixel convolution operation and in matlab code to rotate the instruction kernel.
(15 marks)
Table 1: The matrix of image A and the kernel
1 2 3 4 5 1 136 11 139 65 97
2 42 37 40 140 83
3 90 19 118 53 49
4 86 145 18 48 143
5 54 61 99 19 129
Matrix Image A
8 1 6 3 5 7 4 9 2
Kernel
Question 5
(a) Give the definition of the following term in an image processing perspective:
i. Translation ii. Rotation iii. Scaling
(6 marks)
(b) A point (P) of an object is rotate with an angle of as shown in Figure 5.
Figure 5: Rotation of a point (P) of an object
i. Write the Rotation matrix of the above operation. Then, find the transformed point of an object, P’, after a point P= (8, 2) has rotated in angle of 60˚ as shown in in Figure 5.
(5 marks)
ii. Then the point P’ (b(i)) is translated by (-3, 4) to PT. Write the translation matrix of the operation and find the point Pt.
(5 marks)
iii. Finally the point Pt is scaled by a scale factor (2, 0.5) to PS. Write the scale matrix of the operation and find the point PS.
(5 marks)
(c) Using homogenous composite transformation matrix, get the final point of P (8, 2) if the point is rotated by 60 degrees, then translate by (-3, 4), and finally scaled by the scale factor (2, 0.5). Compare your finding with the result obtained in (b(iii)).
(9 marks)
END OF QUESTIONS